Colorado Becomes First State in US to Regulate AI for Bias

Lawmakers in Colorado have passed the nation’s first bill aimed at protecting people against AI algorithmic discrimination in critical areas like housing, employment, and healthcare, according to a People of Color in Tech article by Samara Linton. This comes just weeks after the U.S. Department of Housing and Urban Development issued a warning that AI-powered software used to select tenants for rental housing and target housing advertisements can introduce bias and potential discrimination into the process. 

Under SB205, developers, deployers, and businesses using “high-risk artificial intelligence systems — defined as those that “can potentially lead to differential treatment based on protected classifications such as age, disability, race, religion, or sex” — will be required to notify the Colorado attorney general of discriminatory risks and outcomes, have risk management strategies in place, and be transparent with the public about AI’s purpose and the types of decisions it’s being used for.

Linton reports that the legislation faced pushback from the tech industry, which argues a state-by-state approach to AI regulation could stifle innovation  — the same argument that led Connecticut Governor  Ned Lamont to veto a similar bill earlier this month. “Other states are closely watching the outcome of Colorado’s legislative efforts. More than 40 states, including California, are considering some 400 AI-related bills, many of which address discrimination issues,” Linton writes.